Anthropic Code Execution vs Vercel Sandbox

Anthropic Code Execution costs $0.05 per vCPU-hour and Vercel Sandbox costs $0.128 per vCPU-hour. Anthropic Code Execution runs containers. Vercel Sandbox runs Firecracker microVMs and limits a session to 24 hours.

Updated

Anthropic Code Execution

Isolation
Container
Free tier
1,550 free hours per month
Max session
Not listed

Vercel Sandbox

Isolation
Firecracker microVM
Free tier
5 CPU-hours every month
Max session
24 hours

What is different

7 fields differ. The rest are the same or not listed.

Anthropic Code Execution
Vercel Sandbox
vCPU, per hour
Anthropic Code Execution: $0.05
Vercel Sandbox: $0.128
RAM, per GiB-hour
Anthropic Code Execution: Included
Vercel Sandbox: $0.0212
Free tier
Anthropic Code Execution: 1,550 free hours per month
Vercel Sandbox: 5 CPU-hours every month
Billing
Anthropic Code Execution: Bundled machine time
Vercel Sandbox: Active CPU time only
Largest sandbox
Anthropic Code Execution: 1 vCPU · 5 GiB
Vercel Sandbox: 8 vCPU · 16 GiB
Isolation
Anthropic Code Execution: Container
Vercel Sandbox: Firecracker microVM
SDKs
Anthropic Code Execution: Python, TypeScript, Go, Java, Ruby
Vercel Sandbox: Python, JavaScript / TypeScript

Full comparison

18 fields, at each provider's list price

Pricing

vCPU, per hour
Anthropic Code Execution: $0.05
Vercel Sandbox: $0.128
RAM, per GiB-hour
Anthropic Code Execution: Included
Vercel Sandbox: $0.0212
Storage
Anthropic Code Execution: Not listed
Vercel Sandbox: On Pro, snapshots $0.08 per GB-month and drives $0.05 per GB-month (beta).
Free tier
Anthropic Code Execution: 1,550 free hours per month
Vercel Sandbox: 5 CPU-hours every month
Paid plan from
Anthropic Code Execution: Not listed
Vercel Sandbox: $20 / month
Billing
Anthropic Code Execution: Bundled machine time
Vercel Sandbox: Active CPU time only

Limits

Max session
Anthropic Code Execution: Not listed
Vercel Sandbox: 24 hours
Idle stop
Anthropic Code Execution: Not listed
Vercel Sandbox: Not listed
Largest sandbox
Anthropic Code Execution: 1 vCPU · 5 GiB
Vercel Sandbox: 8 vCPU · 16 GiB
Concurrent sandboxes
Anthropic Code Execution: Not listed
Vercel Sandbox: Up to 10,000

Platform

Isolation
Anthropic Code Execution: Container
Vercel Sandbox: Firecracker microVM
GPU
Anthropic Code Execution: Not listed
Vercel Sandbox: Not listed
Snapshots
Anthropic Code Execution: Not listed
Vercel Sandbox:
Pause and resume
Anthropic Code Execution:
Vercel Sandbox:

Developer

SDKs
Anthropic Code Execution: Python, TypeScript, Go, Java, Ruby
Vercel Sandbox: Python, JavaScript / TypeScript
Integrations
Anthropic Code Execution: Not listed
Vercel Sandbox: Claude Agent SDK, LangChain, OpenAI Agents SDK, Vercel AI SDK, Claude Code

Trust

License
Anthropic Code Execution: Not listed
Vercel Sandbox: Not listed
BYOC
Anthropic Code Execution: Not listed
Vercel Sandbox: Not listed

Hourly cost

One sandbox running for one hour

Anthropic Code Execution
Vercel Sandbox
1 vCPU · 2 GiB
Anthropic Code Execution: $0.050
Vercel Sandbox: $0.170
2 vCPU · 4 GiB
Anthropic Code Execution: $0.100
Vercel Sandbox: $0.341
4 vCPU · 8 GiB
Anthropic Code Execution: $0.200
Vercel Sandbox: $0.682

Compute only, at list rates.

Open the calculator

Which one fits

Based on the fields above

Anthropic Code Execution fits when

  • You want the lower list price: $0.100 against $0.341 an hour for 2 vCPU and 4 GiB.

Vercel Sandbox fits when

  • You want each sandbox in its own microVM.
  • You use the Claude Agent SDK, LangChain, the OpenAI Agents SDK, the Vercel AI SDK or Claude Code.
  • You need snapshots.
  • Your code waits a lot: CPU is billed only while it is in use.

More comparisons

See all comparisons

Related